Closes #5413: bug in pd.array when dtype is ak.bigint

arkouda/numpy/pdarraycreation.py

@@ -329,7 +329,7 @@ def array(
     # If a is not already a numpy.ndarray, convert it
     if not isinstance(a, np.ndarray):
         try:
-            if dtype is not None and dtype != bigint:
+            if dtype is not None and dtype not in (bigint, "bigint"):
                 # if the user specified dtype, use that dtype
                 a = np.array(a, dtype=dtype)
             elif (

arkouda/pandas/extension/_arkouda_array.py

@@ -114,9 +114,14 @@ def __init__(
     def _from_sequence(cls, scalars, dtype=None, copy=False):
         from arkouda.numpy.pdarraycreation import array as ak_array
 
+        from ._dtypes import ArkoudaBigintDtype
+
         # If pandas passes our own EA dtype, ignore it and infer from data
         if isinstance(dtype, _ArkoudaBaseDtype):
-            dtype = dtype.numpy_dtype
+            if isinstance(dtype, ArkoudaBigintDtype) or getattr(dtype, "name", None) == "bigint":
+                dtype = "bigint"
+            else:
+                dtype = dtype.numpy_dtype
 
         if dtype is not None and hasattr(dtype, "numpy_dtype"):
             dtype = dtype.numpy_dtype

tests/pandas/extension/arkouda_array_extension.py

@@ -389,6 +389,16 @@ def test_copy_default_behaves_like_deep_true(self, ea):
         # Values preserved
         np.testing.assert_array_equal(default_copy.to_numpy(), ea.to_numpy())
 
+    def test_pd_array_construct_from_ak_bigint_pdarray_explicit_dtype_does_not_cast_to_object(self):
+        # Regression: used to try cast<bigint,object,1> and fail on server
+        a = ak.array(np.array([0, -1, 2]), dtype="bigint")
+
+        b = pd.array(a, dtype="ak.bigint")
+
+        # Should be Arkouda-backed and preserve values
+        assert str(b.dtype) in ("ak_bigint", "ak.bigint", "bigint")  # tolerate aliasing
+        np.testing.assert_array_equal(np.asarray(b), np.array([0, -1, 2], dtype=object))
+
 
 class TestArkoudaArrayAsType:
     def test_arkouda_array_astype_object_returns_numpy_object_array(self):